Date: Wed, 28 Oct 2020 20:07:20 -0700 From: David Christensen <dpchrist@holgerdanske.com> To: freebsd-questions@freebsd.org Subject: Re: vf1 kernel pid 609 (ntpd), jid 0, uid 0, was killed; out of swap space Message-ID: <c06f93de-6a01-b262-806b-768a80ad5647@holgerdanske.com> In-Reply-To: <73c9f204-8ef5-0171-bb3b-c3c34e165740@holgerdanske.com> References: <73c9f204-8ef5-0171-bb3b-c3c34e165740@holgerdanske.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2020-10-28 19:05, David Christensen wrote: > 2020-10-28 17:36:43 toor@vf1 ~/src/perl/Dpchrist-Perl-b0_01068002 > # time for i in $( seq 1 100 ) ; do make test ; done After ~140 minutes, the tests are still running, but visibly slower. Logging in via SSH is sluggish. There no new console error messages. Status is as follows: 2020-10-28 19:55:13 toor@vf1 ~ # swapctl -l Device: 1024-blocks Used: /dev/mirror/swap 1048572 139504 2020-10-28 19:55:58 toor@vf1 ~ # top -w -d 1 | head -n 20 last pid: 60001; load averages: 1.10, 1.10, 1.32 up 0+05:28:53 19:56:24 48 processes: 2 running, 46 sleeping CPU: 17.3% user, 0.0% nice, 24.7% system, 2.7% interrupt, 55.4% idle Mem: 22M Active, 1360K Inact, 1136K Laundry, 926M Wired, 4788K Free ARC: 823M Total, 163M MFU, 634M MRU, 153K Anon, 3289K Header, 22M Other 40M Compressed, 760M Uncompressed, 18.97:1 Ratio Swap: 1024M Total, 168M Used, 856M Free, 16% Inuse PID USERNAME THR PRI NICE SIZE RES SWAP STATE C TIME WCPU COMMAND 59998 root 1 22 0 32M 6568K 0 swread 1 0:01 2.98% perl 59999 root 1 25 0 33M 9600K 0 CPU1 1 0:01 2.98% perl 59995 root 1 20 0 42M 6776K 0 swread 1 0:01 0.98% perl 58505 root 1 20 0 25M 1492K 0 select 1 0:11 0.00% perl 1730 root 1 20 0 20M 260K 0 select 1 0:09 0.00% sshd 651 root 1 20 0 17M 276K 0 select 0 0:08 0.00% sendmail 59152 root 1 20 0 25M 1360K 0 select 0 0:07 0.00% perl 4838 root 1 20 0 20M 204K 0 select 0 0:05 0.00% sshd 4834 root 1 20 0 20M 172K 0 select 0 0:05 0.00% sshd 4748 root 1 20 0 20M 148K 0 select 1 0:04 0.00% sshd 5213 root 1 20 0 13M 312K 0 select 0 0:04 0.00% top 2020-10-28 19:56:24 toor@vf1 ~ # top -m io -w -d 1 | head -n 20 last pid: 60004; load averages: 0.93, 1.06, 1.30 up 0+05:28:59 19:56:30 48 processes: 3 running, 45 sleeping CPU: 17.3% user, 0.0% nice, 24.7% system, 2.7% interrupt, 55.4% idle Mem: 19M Active, 1572K Inact, 1692K Laundry, 926M Wired, 6920K Free ARC: 823M Total, 163M MFU, 634M MRU, 128K Anon, 3288K Header, 22M Other 40M Compressed, 760M Uncompressed, 18.97:1 Ratio Swap: 1024M Total, 151M Used, 873M Free, 14% Inuse PID USERNAME VCSW IVCSW READ WRITE FAULT TOTAL PERCENT COMMAND 59999 root 3162 1053 395 2 2952 3349 0.26% perl 59995 root 9057 1779 401 1 8750 9152 0.70% perl 59964 root 9645 1742 375 3 9341 9719 0.75% perl 60002 root 152 245 134 0 158 292 0.02% perl 58505 root 218567 26744 281 0 202628 202909 15.57% perl 1730 root 192785 13715 38 0 178942 178980 13.74% sshd 651 root 124736 6183 1050 1 153814 154865 11.89% sendmail 59152 root 143127 18210 253 0 132596 132849 10.20% perl 4838 root 68504 5564 39 0 32530 32569 2.50% sshd 4834 root 66724 5295 126 0 29236 29362 2.25% sshd 4748 root 60897 4640 113 0 25122 25235 1.94% sshd 2020-10-28 19:56:30 toor@vf1 ~ # zpool list vf1zpool1 NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT vf1zpool1 960M 103M 857M - - 47% 10% 1.00x ONLINE - 2020-10-28 19:56:52 toor@vf1 ~ # zpool iostat -v vf1zpool1 capacity operations bandwidth pool alloc free read write read write --------------- ----- ----- ----- ----- ----- ----- vf1zpool1 103M 857M 2 37 17.4K 430K gpt/vf1zpool1 103M 857M 2 37 17.4K 430K --------------- ----- ----- ----- ----- ----- ----- sendmail(8) seems to be using more I/O than everything else. How do I determine what I/O sendmail(8) is doing, and why? The top(1) load averages are now much lower (1.10 vs. 3.80). How do I determine why? David
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?c06f93de-6a01-b262-806b-768a80ad5647>